The Quiet Contract Nobody Voted On
You open a factory in a country you don't fully understand. The government seems stable, the regulations look manageable, and your lawyers have read the bilateral investment treaty between your home state and the host state three times. You sign. You build. Then, five years later, the host government changes its environmental licensing rules, your plant's permit gets revoked, and you are sitting on a nine-figure sunk cost with no domestic court willing to hear your case fairly. Without that treaty, you have almost nothing. With it, you can drag a sovereign government before an international arbitration tribunal and, if you win, collect damages drawn straight from the public treasury: from teachers' salaries, road budgets, hospital procurement lines.
That is the mechanism. Bilateral investment treaties, known as BITs, are agreements between two countries that define how each will treat investors from the other. They have existed in modern form since West Germany signed one with Pakistan in 1959, and there are now well over three thousand in force worldwide. Each one is, at its core, a legal risk-redistribution instrument. Risk that would otherwise sit entirely with the foreign investor gets transferred, in carefully negotiated increments, onto the host state and its citizens.
What a BIT Actually Promises, Clause by Clause
The language in these treaties sounds procedural and dull. It isn't. Four provisions do most of the work.
Fair and equitable treatment obliges the host state to maintain a stable, transparent regulatory environment. Courts in many jurisdictions have interpreted this broadly: if a government changes its tax code in a way that specifically disadvantages a foreign investor, that can constitute a violation even if the law was perfectly legal under domestic rules. A Dutch energy firm operating wind farms in Spain used this clause to seek hundreds of millions of euros in arbitration after Spain reduced its renewable energy subsidies, arguing that the regulatory environment at the time of investment had been materially altered. The claim was not that the law was corrupt. The claim was simply that things had changed.
Most-favored-nation clauses require that the host state give your investors treatment at least as favorable as it gives investors from any other country. This sounds like basic fairness. In practice, it means investors can sometimes import procedural protections from other treaties the host state has signed, shopping across the treaty network for whichever terms suit them best. The network, in this sense, functions less like a web of bilateral promises and more like a vast menu.
Expropriation protections cover not just outright nationalization but also indirect expropriation: regulatory measures that, without formally taking your asset, strip it of economic value. The line between a government legitimately regulating in the public interest and illegally expropriating is genuinely contested, and arbitral panels draw it differently. There is no supreme court to reconcile their rulings.
Finally, investor-state dispute settlement, or ISDS, is the enforcement engine. It lets corporations bypass domestic courts and take claims directly to international arbitration, typically under rules administered by bodies like the International Centre for Settlement of Investment Disputes, housed at the World Bank. Awards are enforceable against state assets in most countries. That enforceability is what gives the whole framework teeth.
The Arithmetic of Shifted Risk: A Worked Scenario
Consider two investors, both building pharmaceutical manufacturing plants in a middle-income country. Call them Marta and Jonas. Marta's home country holds a BIT with the host state. Jonas's does not.
The host government subsequently passes a law requiring all manufacturers to use locally sourced raw materials, raising production costs by roughly thirty percent and making both plants economically marginal. Marta's lawyers file for arbitration, arguing the new law constitutes an indirect expropriation and violates fair and equitable treatment. If the tribunal agrees, the host state pays Marta compensation calibrated to the plant's lost future profits, not just its book value. Jonas gets nothing from international law. He sues in domestic courts, waits years, and likely loses before judges appointed by the very government that passed the law.
One policy. One economic harm. Radically different legal exposure, depending entirely on which passport Marta and Jonas hold.
That asymmetry is not a bug in the system. It is the product. For the host state, the consequences compound quietly and in advance: governments facing potential arbitration claims sometimes shelve regulations they would otherwise have passed without hesitation. Researchers call this the regulatory chill effect, and it is, in this writer's view, a severely underreported cost in the entire debate. The risk has not disappeared. It sits internalized within the state as a kind of permanent actuarial shadow over policymaking, a silent veto held by no elected official.
What People Get Wrong About Who Benefits
The standard critique frames BITs as instruments that wealthy states use to discipline poorer ones. That critique carries real force. Historically, capital-exporting countries in Europe and North America wrote the early treaties, and capital-importing developing countries signed them partly because they believed the treaties would attract foreign investment. The evidence that BITs actually increase investment flows is surprisingly weak. Multiple econometric studies have found no robust causal relationship between treaty signing and increased capital inflows, suggesting that investors care about broader institutional quality, not the treaty text alone. The treaties were sold as a growth instrument; the receipts tell a different story.
But the picture has shifted. Countries like China, India, and Brazil have become major outward investors, and their corporations increasingly use BITs to protect assets in Africa, Southeast Asia, and Latin America. The tool gets picked up by whoever holds it. A Chinese mining company operating in a sub-Saharan country can, depending on applicable treaties, access an arbitration architecture broadly comparable to what a German automaker would use. The original North-South framing has become genuinely obsolete in some corridors of the system.
Ask yourself this: when did you last see a parliamentary debate about the specific ISDS clause in a treaty your government signed a decade ago?
You almost certainly did not. The treaties are negotiated by officials in foreign ministries, ratified with minimal legislative scrutiny in many countries, and the arbitral awards are paid from public budgets. The tobacco company Philip Morris used investor protections under a trade agreement to challenge Australia's plain-packaging laws, a public health measure backed by substantial evidence. Philip Morris eventually lost, but Australia spent years and considerable public funds defending a domestic health policy in a private arbitral forum staffed by lawyers billing by the hour. That is what risk redistribution looks like on the ground: not a line in a treaty, but a health ministry's legal budget being consumed by a dispute that no Australian voter ever anticipated.
The BIT system is not going away. It will evolve through renegotiation, through new multilateral frameworks such as the EU's proposed multilateral investment court, and through countries terminating old treaties and writing new ones with narrower ISDS clauses. Some of that reform is genuinely encouraging; much of it is cosmetic. What will not change is the underlying tension: any legal instrument that makes it cheaper for capital to cross borders also makes it more expensive for governments to govern. That is not a judgment about whether foreign investment is desirable. It is a judgment about who pays the bill when the two interests collide, and the answer, consistently, is the party that was never at the negotiating table.